Epstein-Barr virus (EBV) infection is associated with B-cell lymphomas and oral hairy leukoplakia (particularly in AIDS patients), as well as nasopharyngeal carcinoma (NPC) and gastric cancer. Recent findings from our lab indicate that lytically-infected cells secrete cytokines and other factors that may promote growth of tumor cells in vivo. In addition, increasing evidence suggests that a secreted viral protein, BARF1 (which was previously reported to be a decoy receptor for the macrophage differentiation factor, CSF-1), plays a critical role in the etiology of EBV+ epithelial cancers, and may also contribute to Burkitt lymphomas (BL). Although the switch from latent to lytic infection is thought to be mediated by the BZLF1 (Z) and BRLF1(R) proteins, we have discovered that expression of another viral protein, BRRF1 (Na), is sufficient to induce the lytic form of infection in latently infected NPC and gastric cancer cells. In addition, our new preliminary data suggest that BARF1 inhibits lytic viral gene expression in epithelial cells, while CSF-1 signaling enhances lytic gene expression. Therefore, BARF1 may inhibit lytic reactivation by blocking CSF-1-mediated lytic induction. Thus, regulation of Na transcription, as well as Na and BARF1 function, must play key roles in determining if EBV infection in epithelial cells is latent or lytic. To study the roles of specific EBV proteins such as Na and BARF1 in viral pathogenesis, we have developed a new mouse model (engrafted with a reconstituted human immune system in the presence or absence of a human thymus) that supports both latent and lytic infection, as well as lymphomas, in EBV-infected mice. Based upon our exciting preliminary data, we propose the following aims. In Aim 1, we will examine the in vitro phenotype of a Na-mutant virus in B cells versus epithelial cells, and elucidate the mechanism(s) by which Na mediates viral reactivation in epithelial cells. In Aim 2, we will define the viral and cellular factors that regulate Na transcription in the presence and absence of promoter methylation, and determine if Na is an IE or early gene in different cell types. In Aim 3, we will explore how CSF-1 and BARF1 regulate viral reactivation and growth of EBV-infected cells in vitro, and determine if BARF1 promotes the growth of EBV- positive epithelial tumors in SCID mice. In Aim 4, we will define the in vivo phenotypes of the Na and BARF1 mutant viruses in our new mouse model containing an engrafted human immune system. The proposed studies should yield important insights into how the switch from latent to lytic infection is regulated in B cells versus epithelial cells, and define the multiple roles that lytic EB proteins play in EBV-associated diseases.